Anammox and denitrification carried out by microorganisms in marine oxygen minimum zones (OMZs) contribute 20-40% of oceanic nitrogen (N)-loss. While these processes rely on sinking particles to provide organic matter and ammonium, the links between particle properties, microbial communities and N-loss rates were previously unknown. The aim of my project was therefore to quantify potential N-loss rates in individual particles, characterize the responsible microbial communities and test which properties of the sinking particles control N-loss in the Peruvian OMZ. Bulk water anammox rates were significantly correlated with in situ abundances of particles.
A mechanistic model revealed that especially smaller particles (128-256 µm) played a dominant role in transporting and releasing ammonium and the likelihood of an anammox cell to encounter the ammonium hotspot around a small particle was highest. As a result, the correlation of anammox rates with the abundance of small particles could be applied successfully to carry out regional N-loss estimates (Karthäuser and Ahmerkamp et al. 2021, Nat. Comm.). Variable rates of denitrification were associated with ~80% of the individual particles, which could explain the previously observed variability in denitrification bulk water rates. A metagenomic characterization of denitrifying microbial communities on individual particles revealed a previously overlooked heterogeneity that is introduced due to the individual particle’s colonization history and the formation of denitrifying consortia on particles. Combining measurements of sinking velocities, carbon contents and various visualization methods further revealed a heterogeneous composition of materials in the particles. As larger particles were often more porous, denitrification rates were only loosely correlated with particle size. The total contribution of different particle types to N-loss was mainly controlled by their sinking velocities and resulting residence times in the OMZ. These properties might therefore be relevant in predicting denitrification in future modeling studies and it is likely that future estimates could benefit from linking visual properties captured by in situ imaging techniques with the physical properties of particles. Taken together, these results contribute to the mechanistic understanding of both N-loss processes in marine OMZs and can be used for regional estimates based on particle abundance profiles.